
import sqlite3
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
import pandas as pd
from pandas_visual_analysis import VisualAnalysis
from sklearn.neighbors import KNeighborsClassifier
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier
from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score, confusion_matrix
from sklearn.discriminant_analysis import LinearDiscriminantAnalysis, QuadraticDiscriminantAnalysis
from sklearn.neighbors import KNeighborsClassifier
from sklearn.metrics import confusion_matrix, classification_report
from sklearn.tree import DecisionTreeRegressor
from sklearn.svm import SVR
from sklearn.ensemble import RandomForestRegressor
from sklearn.preprocessing import StandardScaler
from sklearn.model_selection import GridSearchCV, cross_val_score
from sklearn.model_selection import train_test_split
from sklearn.linear_model import Perceptron, LogisticRegression
from sklearn.model_selection import train_test_split, cross_val_score
from sklearn.metrics import accuracy_score, ConfusionMatrixDisplay
from sklearn.preprocessing import LabelEncoder
from sklearn.linear_model import LinearRegression, Lasso, Ridge
from sklearn.model_selection import GridSearchCV
from sklearn.metrics import mean_squared_error
from scipy.stats import shapiro, boxcox_normmax, boxcox
from statsmodels.graphics.gofplots import qqplot
from sklearn.preprocessing import StandardScaler
from sklearn.preprocessing import MinMaxScaler
import warnings
warnings.filterwarnings("ignore")


df = pd.read_excel("./session_7_dataset.xlsx")


df.drop("Unnamed: 0", axis=1, inplace=True)

df


print("Type of data:", type(df))
print("\nDimensions: \nNumber of rows:",df.shape[0], "\nNumber of columns:",df.shape[1])
df

Type of data: <class 'pandas.core.frame.DataFrame'>

Dimensions: 
Number of rows: 17857 
Number of columns: 25


#Basic Information about dataset
df.info()

df.describe()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 17857 entries, 0 to 17856
Data columns (total 25 columns):
 #   Column                                           Non-Null Count  Dtype  
---  ------                                           --------------  -----  
 0   inm_floor                                        14067 non-null  float64
 1   inm_size                                         16090 non-null  float64
 2   inm_price                                        17857 non-null  int64  
 3   inm_longitude                                    17857 non-null  float64
 4   inm_latitude                                     17857 non-null  float64
 5   inm_barrio                                       16063 non-null  object 
 6   inm_distrito                                     16110 non-null  object 
 7   his_price                                        17857 non-null  int64  
 8   his_quarterly_variation                          17857 non-null  float64
 9   his_annual_variation                             17857 non-null  float64
 10  his_monthly_variation                            17857 non-null  float64
 11  dem_Indice_de_reemplazo_de_la_poblacion_activa   17857 non-null  float64
 12  dem_Indice_de_juventud                           17857 non-null  float64
 13  dem_Indice_de_estructura_de_la_poblacion_activa  17857 non-null  float64
 14  dem_Indice_de_dependencia                        17857 non-null  float64
 15  dem_TasaDeParo                                   17857 non-null  float64
 16  dem_TamanoMedioDelHogar                          17857 non-null  float64
 17  dem_PropSinEstudiosUniversitarios                17857 non-null  float64
 18  dem_PropSinEstudios                              17857 non-null  float64
 19  dem_Proporcion_de_nacidos_fuera_de_Espana        17857 non-null  float64
 20  dem_PropConEstudiosUniversitarios                16132 non-null  float64
 21  dem_PobTotal                                     17857 non-null  int64  
 22  dem_NumViviendas                                 17857 non-null  int64  
 23  dem_EdadMedia                                    16067 non-null  float64
 24  dem_Densidad_(Habit/Ha)                          17857 non-null  float64
dtypes: float64(19), int64(4), object(2)
memory usage: 3.4+ MB


#Check for non-numerical columns
df.select_dtypes(exclude=np.number)


print("Number of neighborhoods: ", len(df.inm_barrio.unique()))
print("Number of districts: ", len(df.inm_distrito.unique()))

Number of neighborhoods:  122
Number of districts:  22


VisualAnalysis(df)

VBox(children=(ToggleButtons(_dom_classes=('layout-de7532dc8f6d4ff99419148606e75d25',), description='Selection…


#Check for duplicates
print("Number of Duplicates", df.duplicated().sum())

Number of Duplicates 6


#Drop duplicates - its number is not that significant for the overall dataset
df.drop_duplicates(inplace=True)


from pandas.api.types import is_numeric_dtype

#Missing values for Each Category
missing_columns_num=[]
missing_columns_non_num=[]
for col in df.columns:
    if df[col].isnull().values.any():
        print(col)
        if is_numeric_dtype(df[col]):
            missing_columns_num.append(col)
        else:
            missing_columns_non_num.append(col)
        missing_count = df[col].isnull().sum()
        print("Missing Values: ", missing_count,  "({:.2%})\n".format(missing_count/ df.shape[0]))

sns.heatmap(df.isnull(), yticklabels=False, cbar=False, cmap="viridis")

inm_floor
Missing Values:  3790 (21.23%)

inm_size
Missing Values:  1766 (9.89%)

inm_barrio
Missing Values:  1794 (10.05%)

inm_distrito
Missing Values:  1747 (9.79%)

dem_PropConEstudiosUniversitarios
Missing Values:  1725 (9.66%)

dem_EdadMedia
Missing Values:  1790 (10.03%)

<AxesSubplot:>


#Fill numerical columns with mean 
for col in missing_columns_num:
    mean_value=df[col].mean()
    df[col].fillna(value=mean_value, inplace=True)
    
#Fill numerical columns with mean
for col in missing_columns_non_num:
    df[col].fillna(method="ffill", inplace=True)


# Univariate Analysis for Numerical Columns
numeric_cols = df.select_dtypes(include=np.number)

# plot boxplot for each numeric column
for col in numeric_cols.columns.tolist():
    sns.boxplot(df[col])
    plt.title(col)
    plt.show()

df


columns=numeric_cols.columns
df_checkpoint=df.copy()

def remove_outliers(dataset, column):
    q1, q3 = np.percentile(df[column], [25, 75])
    iqr = q3 - q1
    lower_bound = q1 - (1.5 * iqr)
    upper_bound = q3 + (1.5 * iqr)
    dataset = dataset[(dataset[column] >= lower_bound) & (dataset[column] <= upper_bound)]
    return dataset
        

for col in columns:
    df=remove_outliers(df, col)


# plot histogram for each numeric column
for col in numeric_cols.columns.tolist():
    sns.histplot(df[col])
    plt.title(col)
    plt.show()


# Univariate Analysis for Non-Numerical Columns
plt.tight_layout()

non_numeric_cols = df.select_dtypes(exclude=np.number)
#plot countplot for each non-numeric column
for col in non_numeric_cols.columns.tolist():
    fig, ax = plt.subplots(figsize=(20, 7))
    sns.countplot(df[col])
    
    plt.title(col)
    plt.xticks(rotation=45)
    plt.show()

<Figure size 432x288 with 0 Axes>


sns.pairplot(df)
plt.show()


sns.set_style("darkgrid")
def plot_correlation(df, dependent_var):
    """
    Plot correlation between dependent variable and each independent variable
    along with mean price for each category if dependent variable is categorical.
    """
    for col in df.columns:
        if col == dependent_var:
            continue
        if df[col].dtype == 'object':
            # Create a bar plot showing mean price for each category
            fig, ax = plt.subplots(figsize=(20, 7))
            plt.xticks(rotation=45)
            plt.ylabel('mean_price')
            sns.barplot(x=col, y="inm_price", data=df)
            plt.show()
        else:
            # Create a scatter plot showing correlation between independent variable and price
            sns.scatterplot(x=col, y="inm_price", data=df)
            plt.show()

plot_correlation(df, "inm_price")


# Correlation Analysis
corr = df.select_dtypes(include=np.number).corr()
f, ax = plt.subplots(figsize=(22, 22))
sns.heatmap(corr, vmax=.8, square=True)
plt.title('Correlation Matrix')
plt.show()


def detect_correlation(data, threshold):
    correlation_set = set()  
    correlation_matrix = data.corr()
    for i in range(len(correlation_matrix.columns)):
        for j in range(i):
            if abs(correlation_matrix.iloc[i, j]) > threshold: 
                correlation_set.add(correlation_matrix.columns[i])
    if len(correlation_set) == 0:
        print("No correlated variables detected")
    else: 
        return correlation_set

detect_correlation(df, 0.8)

{'dem_Indice_de_juventud',
 'dem_NumViviendas',
 'dem_PropConEstudiosUniversitarios',
 'dem_PropSinEstudios',
 'dem_PropSinEstudiosUniversitarios',
 'dem_TamanoMedioDelHogar',
 'dem_TasaDeParo'}


df.drop(["dem_NumViviendas", 
         "dem_PropConEstudiosUniversitarios", 
         "dem_PropSinEstudios", 
         "dem_PropSinEstudiosUniversitarios",
         "dem_TamanoMedioDelHogar",
         "dem_TasaDeParo",
         "dem_Indice_de_juventud"
        ], axis=1, inplace=True)


categorical_variables = ["inm_barrio", "inm_distrito"]
for col in categorical_variables:
    print(df[col].value_counts())

Universidad      490
Goya             426
Justicia         337
Lista            305
Palacio          303
                ... 
Corralejos         1
Valdebernardo      1
Arcos              1
La Paz             1
Rosas              1
Name: inm_barrio, Length: 116, dtype: int64
Centro                   1471
Salamanca                1456
Tetuán                    998
Chamberí                  877
Carabanchel               859
Arganzuela                692
Chamartín                 624
Puente de Vallecas        524
Ciudad Lineal             506
Moncloa - Aravaca         417
Usera                     365
Latina                    309
Retiro                    277
Hortaleza                 216
Villaverde                177
Fuencarral - El Pardo      97
San Blas - Canillejas      86
Moratalaz                  71
Villa de Vallecas          57
Vicálvaro                  32
Barajas                     7
Name: inm_distrito, dtype: int64


df_checkpoint1=df.copy()
def remove_small_categories(df, categorical_column):
    # Calculate value counts for the categorical column
    counts = df[categorical_column].value_counts()
    df = df[~df[categorical_column].isin(counts[counts < 15].index)]
    return df

# for cat in categorical_variables:
df=remove_small_categories(df, "inm_barrio")


categorical_variables = ["inm_barrio", "inm_distrito"]
for col in categorical_variables:
    print(df[col].value_counts())

Universidad                     490
Goya                            426
Justicia                        337
Lista                           305
Palacio                         303
                               ... 
Marroquina                       24
Hellín                           20
Piovera                          19
Campamento                       16
Casco histórico de Vicálvaro     16
Name: inm_barrio, Length: 76, dtype: int64
Centro                   1449
Salamanca                1423
Tetuán                    975
Chamberí                  868
Carabanchel               839
Arganzuela                676
Chamartín                 618
Puente de Vallecas        521
Ciudad Lineal             494
Moncloa - Aravaca         397
Usera                     359
Latina                    303
Retiro                    272
Hortaleza                 209
Villaverde                174
Fuencarral - El Pardo      93
San Blas - Canillejas      83
Moratalaz                  68
Villa de Vallecas          55
Vicálvaro                  30
Barajas                     7
Name: inm_distrito, dtype: int64


numeric_cols = df.select_dtypes(include=np.number)

# plot histogram for each numeric column
for col in numeric_cols.columns.tolist():
    #sns.histplot(df[col])
    sns.distplot(df[col].values, color='g')
    plt.title(col)
    plt.show()


col_names=numeric_cols.columns

scaler = StandardScaler()

for col in col_names:
    df[col] = scaler.fit_transform(df[col].values.reshape(-1,1))
    
df.head()


numeric_cols = df.select_dtypes(include=np.number)

# plot histogram AFTER transforamtion
for col in numeric_cols.columns.tolist():
    #sns.histplot(df[col])
    sns.distplot(df[col].values, color='g')
    plt.title(col)
    plt.show()


df_checkpoint1=df.copy()


#Dummy Variables
distrito_dummy= pd.get_dummies(df['inm_distrito'], prefix='distrito_').reset_index(drop=True)

# create dummy variables for the 'size' column
bario_dummy = pd.get_dummies(df['inm_barrio'], prefix='bario_').reset_index(drop=True)


# concatenate the original dataframe with the dummy variables
df1 = pd.concat([df.reset_index(drop=True), bario_dummy, distrito_dummy], axis=1)

df1.drop(["inm_barrio", "inm_distrito"], axis=1, inplace=True)

df1


df=df1.copy()


from sklearn.model_selection import train_test_split
y=df.inm_price 
X=df.drop("inm_price", axis=1)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)


# train a linear regression model
linear_reg = LinearRegression()
linear_reg.fit(X_train, y_train)
y_pred_linear = linear_reg.predict(X_test)
mse_linear = mean_squared_error(y_test, y_pred_linear)
print("Linear Regression MSE: ", mse_linear)

# train a Lasso regression model with cross-validation
lasso_reg = Lasso()
parameters = {'alpha': np.logspace(-4, 4, 9)}
grid_search_lasso = GridSearchCV(lasso_reg, parameters, scoring='neg_mean_squared_error', cv=5)
grid_search_lasso.fit(X_train, y_train)
best_lasso_reg = grid_search_lasso.best_estimator_
y_pred_lasso = best_lasso_reg.predict(X_test)
mse_lasso = mean_squared_error(y_test, y_pred_lasso)
print("\n\nLasso Regression MSE: ", mse_lasso)
print("Lasso Regression Best Coefficient: ", best_lasso_reg)

# train a Ridge regression model with cross-validation
ridge_reg = Ridge()
parameters = {'alpha': np.logspace(-4, 4, 9)}
grid_search_ridge = GridSearchCV(ridge_reg, parameters, scoring='neg_mean_squared_error', cv=5)
grid_search_ridge.fit(X_train, y_train)
best_ridge_reg = grid_search_ridge.best_estimator_
y_pred_ridge = best_ridge_reg.predict(X_test)
mse_ridge = mean_squared_error(y_test, y_pred_ridge)
print("\n\nRidge Regression MSE: ", mse_ridge)
print("Ridge Regression Best Coefficient: ", best_ridge_reg)

# Decision Tree
from sklearn.tree import DecisionTreeRegressor

# Creating the Decision Tree model
dt_reg = DecisionTreeRegressor()

# Defining the parameters to be tuned
parameters = {'max_depth': [2, 4, 6, 8], 'min_samples_leaf': [1, 2, 4, 8], 'min_samples_split': [2, 4, 8, 16]}

# Performing GridSearchCV
grid_search_dt = GridSearchCV(dt_reg, parameters, scoring='neg_mean_squared_error', cv=5)
grid_search_dt.fit(X_train, y_train)
best_dt_reg = grid_search_dt.best_estimator_

# Making the predictions on the test set
y_pred_dt = best_dt_reg.predict(X_test)

# Calculating the mean squared error
mse_dt = mean_squared_error(y_test, y_pred_dt)

# Printing the results
print("\n\nDecision Tree Regression MSE: ", mse_dt)
print("Decision Tree Regression Best Coefficient: ", best_dt_reg)

import xgboost as xgb

# Creating the XGBoost model
xgb_reg = xgb.XGBRegressor()

# Defining the parameters to be tuned
parameters = {'learning_rate': [0.01, 0.1, 0.2, 0.3], 'max_depth': [2, 4, 6, 8], 'n_estimators': [100, 200, 300, 400]}

# Performing GridSearchCV
grid_search_xgb = GridSearchCV(xgb_reg, parameters, scoring='neg_mean_squared_error', cv=5)
grid_search_xgb.fit(X_train, y_train)
best_xgb_reg = grid_search_xgb.best_estimator_

# Making the predictions on the test set
y_pred_xgb = best_xgb_reg.predict(X_test)

# Calculating the mean squared error
mse_xgb = mean_squared_error(y_test, y_pred_xgb)

# Printing the results
print("\n\nXGBoost Regression MSE: ", mse_xgb)
print("XGBoost Regression Best Coefficient: ", best_xgb_reg)

Linear Regression MSE:  0.24155706810315067


Lasso Regression MSE:  0.24032651947533892
Ridge Lasso Best Coefficient:  Lasso(alpha=0.0001)


Ridge Regression MSE:  0.23997332256114737
Ridge Regression Best Coefficient:  Ridge(alpha=10.0)


Decision Tree Regression MSE:  0.19428103180053047
Decision Tree Regression Best Coefficient:  DecisionTreeRegressor(max_depth=8, min_samples_split=8)


XGBoost Regression MSE:  0.15528672614791553
XGBoost Regression Best Coefficient:  XGBRegressor(base_score=0.5, booster='gbtree', callbacks=None,
             colsample_bylevel=1, colsample_bynode=1, colsample_bytree=1,
             early_stopping_rounds=None, enable_categorical=False,
             eval_metric=None, feature_types=None, gamma=0, gpu_id=-1,
             grow_policy='depthwise', importance_type=None,
             interaction_constraints='', learning_rate=0.1, max_bin=256,
             max_cat_threshold=64, max_cat_to_onehot=4, max_delta_step=0,
             max_depth=8, max_leaves=0, min_child_weight=1, missing=nan,
             monotone_constraints='()', n_estimators=200, n_jobs=0,
             num_parallel_tree=1, predictor='auto', random_state=0, ...)


from sklearn import tree
fig = plt.figure(figsize=(15,15))
_ = tree.plot_tree(best_dt_reg, feature_names=df.columns, filled=True, max_depth=3, fontsize=10)
plt.show()


df2=df_checkpoint1.copy()


#Dummy Variables
distrito_dummy_class = pd.get_dummies(df2['inm_distrito'], prefix='distrito_').reset_index(drop=True)

# concatenate the original dataframe with the dummy variables
df3 = pd.concat([df2.reset_index(drop=True), distrito_dummy_class], axis=1)

df3.drop(["inm_distrito"], axis=1, inplace=True)


le = LabelEncoder()
target_encoded = le.fit_transform(df3.inm_barrio)


X_train, X_test, y_train, y_test = train_test_split(df3.drop(["inm_barrio"], axis=1), target_encoded, test_size=0.2)


from sklearn.metrics import accuracy_score, f1_score, confusion_matrix

# create Perceptron model
perceptron = Perceptron(random_state=42)
perceptron_scores = cross_val_score(perceptron, df3.drop(["inm_barrio"], axis=1), target_encoded, cv=5)
print("Cross-Validation Results\n")
print("Perceptron accuracy (cross-validation):", perceptron_scores.mean())

# create Logistic Regression model
logreg = LogisticRegression(random_state=42, multi_class='ovr')
logreg_scores = cross_val_score(logreg, df3.drop(["inm_barrio"], axis=1), target_encoded, cv=5)
print("Logistic Regression accuracy (cross-validation):", logreg_scores.mean())

# fit the models on the training data
perceptron.fit(X_train, y_train)
logreg.fit(X_train, y_train)

# predict the labels for the test data
y_pred_perceptron = perceptron.predict(X_test)
y_pred_logreg = logreg.predict(X_test)

# calculate and print the accuracies of the models
acc_perceptron = accuracy_score(y_test, y_pred_perceptron)
acc_logreg = accuracy_score(y_test, y_pred_logreg)

# calculate and print the f1 scores of the models
f1_perceptron = f1_score(y_test, y_pred_perceptron, average='weighted')
f1_logreg = f1_score(y_test, y_pred_logreg, average='weighted')

# calculate and print the confusion matrix of the models
cm_perceptron = confusion_matrix(y_test, y_pred_perceptron)
cm_logreg = confusion_matrix(y_test, y_pred_logreg)

# calculate and print the confusion matrix of the models
cm_perceptron = confusion_matrix(y_test, y_pred_perceptron)
cm_logreg = confusion_matrix(y_test, y_pred_logreg)
print("\nTest Results\n")
print("Perceptron accuracy (test set):", acc_perceptron)
print("Logistic Regression accuracy (test set):", acc_logreg)

print("\nF1 Scores\n")
print("Perceptron F1 score (test set):", f1_perceptron)
print("Logistic Regression F1 score (test set):", f1_logreg)

# visualize the confusion matrix
fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(20, 10))
sns.heatmap(cm_perceptron, annot=True, cmap='Blues', fmt='g', ax=axes[0])
sns.heatmap(cm_logreg, annot=True, cmap='Blues', fmt='g', ax=axes[1])
axes[0].set_title('Perceptron Confusion Matrix')
axes[1].set_title('Logistic Regression Confusion Matrix')
plt.show()

Cross-Validation Results

Perceptron accuracy (cross-validation): 0.826996218615039
Logistic Regression accuracy (cross-validation): 0.9152631881588864

Test Results

Perceptron accuracy (test set): 0.8693898134140191
Logistic Regression accuracy (test set): 0.9157841654059505

F1 Scores

Perceptron F1 score (test set): 0.8594728749800582
Logistic Regression F1 score (test set): 0.9120210788672336


# Create an LDA model
lda = LinearDiscriminantAnalysis()

# Perform cross-validation to tune the hyperparameters
params = {'solver': ['svd', 'lsqr', 'eigen']}
lda_cv = GridSearchCV(lda, params, cv=5)
lda_cv.fit(X_train, y_train)

# Print the best hyperparameters
print("Best Hyperparameters for LDA: ", lda_cv.best_params_)

# Fit the model using the best hyperparameters
lda = LinearDiscriminantAnalysis(**lda_cv.best_params_)
lda.fit(X_train, y_train)

# Evaluate the model using the testing set
lda_predictions = lda.predict(X_test)
print("LDA Classification Report: ")
print(classification_report(y_test, lda_predictions))

# Plot the confusion matrix
lda_cm = confusion_matrix(y_test, lda_predictions)
plt.figure(figsize=(8,6))
sns.heatmap(lda_cm, annot=True, cmap='Blues', fmt='g')
plt.xlabel('Predicted labels')
plt.ylabel('True labels')
plt.title('LDA Confusion Matrix')
plt.show()

# Create a QDA model
qda = QuadraticDiscriminantAnalysis()

# Perform cross-validation to tune the hyperparameters
params = {'reg_param': [0.0, 0.1, 0.5, 1.0]}
qda_cv = GridSearchCV(qda, params, cv=5)
qda_cv.fit(X_train, y_train)

# Print the best hyperparameters
print("Best Hyperparameters for QDA: ", qda_cv.best_params_)

# Fit the model using the best hyperparameters
qda = QuadraticDiscriminantAnalysis(**qda_cv.best_params_)
qda.fit(X_train, y_train)

# Evaluate the model using the testing set
qda_predictions = qda.predict(X_test)
print("QDA Classification Report: ")
print(classification_report(y_test, qda_predictions))

# Plot the confusion matrix
qda_cm = confusion_matrix(y_test, qda_predictions)
plt.figure(figsize=(8,6))
sns.heatmap(qda_cm, annot=True, cmap='Blues', fmt='g')
plt.xlabel('Predicted labels')
plt.ylabel('True labels')
plt.title('QDA Confusion Matrix')
plt.show()

Best Hyperparameters for LDA:  {'solver': 'svd'}
LDA Classification Report: 
              precision    recall  f1-score   support

           0       1.00      0.85      0.92        26
           1       0.84      0.91      0.87        34
           2       0.64      0.70      0.67        10
           3       0.94      0.85      0.89        40
           4       0.92      0.92      0.92        38
           5       0.85      0.85      0.85        20
           6       1.00      0.95      0.98        22
           7       0.67      0.89      0.77        35
           8       0.84      0.96      0.90        27
           9       0.92      0.94      0.93        36
          10       1.00      0.50      0.67         4
          11       0.79      0.94      0.86        16
          12       0.00      0.00      0.00         1
          13       0.67      1.00      0.80         4
          14       0.82      0.82      0.82        40
          15       0.95      0.95      0.95        19
          16       1.00      0.79      0.88        19
          17       0.81      0.89      0.85        19
          18       0.77      0.94      0.85        18
          19       1.00      0.92      0.96        13
          20       0.91      1.00      0.95        20
          21       0.89      0.93      0.91        27
          22       0.88      0.88      0.88        16
          23       0.84      0.86      0.85        37
          24       0.88      0.88      0.88        32
          25       1.00      0.85      0.92        20
          26       0.00      0.00      0.00         6
          27       0.93      0.88      0.90        16
          28       1.00      1.00      1.00        10
          29       0.96      0.96      0.96        23
          30       0.88      0.86      0.87        43
          31       0.92      0.95      0.93        82
          32       0.96      0.87      0.91        60
          33       1.00      1.00      1.00         3
          34       0.96      1.00      0.98        23
          35       1.00      0.76      0.86        29
          36       0.83      0.90      0.86        21
          37       1.00      0.90      0.95        20
          38       0.97      0.95      0.96        77
          39       0.90      0.87      0.89        54
          40       0.76      0.90      0.83        29
          41       0.96      0.96      0.96        25
          42       0.80      0.80      0.80         5
          43       0.80      0.84      0.82        19
          44       0.93      0.88      0.90        16
          45       0.91      0.95      0.93        41
          46       0.97      0.92      0.95        38
          47       0.92      0.94      0.93        63
          48       0.87      1.00      0.93        27
          49       0.89      0.92      0.91        26
          50       0.79      0.88      0.83        17
          51       0.75      0.86      0.80        14
          52       0.50      0.20      0.29         5
          53       0.95      0.95      0.95        20
          54       0.94      0.84      0.89        19
          55       0.97      0.94      0.96        36
          56       0.85      0.85      0.85        20
          57       0.81      0.81      0.81        26
          58       1.00      0.88      0.94        25
          59       0.92      0.88      0.90        25
          60       0.92      0.92      0.92        39
          61       0.60      1.00      0.75         3
          62       0.00      0.00      0.00         5
          63       0.95      0.92      0.94        39
          64       0.93      0.93      0.93        14
          65       0.80      0.80      0.80        10
          66       0.75      0.86      0.80         7
          67       0.94      0.81      0.87        37
          68       0.91      0.84      0.87        49
          69       0.95      0.95      0.95        78
          70       0.89      0.94      0.92        35
          71       0.93      0.81      0.87        16
          72       0.96      1.00      0.98        46
          73       0.82      1.00      0.90        31
          74       0.93      0.93      0.93        14
          75       0.80      1.00      0.89         4

    accuracy                           0.90      1983
   macro avg       0.85      0.85      0.85      1983
weighted avg       0.90      0.90      0.90      1983

Best Hyperparameters for QDA:  {'reg_param': 1.0}
QDA Classification Report: 
              precision    recall  f1-score   support

           0       1.00      0.88      0.94        26
           1       0.84      0.94      0.89        34
           2       0.75      0.90      0.82        10
           3       0.95      0.88      0.91        40
           4       0.92      0.92      0.92        38
           5       0.86      0.90      0.88        20
           6       1.00      1.00      1.00        22
           7       0.80      0.91      0.85        35
           8       0.84      0.96      0.90        27
           9       0.92      0.94      0.93        36
          10       1.00      0.50      0.67         4
          11       0.88      0.94      0.91        16
          12       1.00      1.00      1.00         1
          13       0.80      1.00      0.89         4
          14       0.82      0.82      0.82        40
          15       0.95      0.95      0.95        19
          16       1.00      0.79      0.88        19
          17       0.86      0.95      0.90        19
          18       0.89      0.94      0.92        18
          19       1.00      0.92      0.96        13
          20       0.87      1.00      0.93        20
          21       0.89      0.93      0.91        27
          22       0.88      0.88      0.88        16
          23       0.89      0.89      0.89        37
          24       0.91      0.91      0.91        32
          25       1.00      0.90      0.95        20
          26       0.00      0.00      0.00         6
          27       0.94      0.94      0.94        16
          28       1.00      1.00      1.00        10
          29       1.00      0.91      0.95        23
          30       0.90      0.86      0.88        43
          31       0.94      0.98      0.96        82
          32       0.96      0.92      0.94        60
          33       1.00      1.00      1.00         3
          34       0.96      0.96      0.96        23
          35       1.00      0.86      0.93        29
          36       0.83      0.95      0.89        21
          37       1.00      0.90      0.95        20
          38       0.97      0.97      0.97        77
          39       0.91      0.94      0.93        54
          40       0.82      0.93      0.87        29
          41       0.96      1.00      0.98        25
          42       1.00      1.00      1.00         5
          43       0.89      0.84      0.86        19
          44       0.94      0.94      0.94        16
          45       0.89      0.98      0.93        41
          46       1.00      0.92      0.96        38
          47       0.92      0.94      0.93        63
          48       0.87      1.00      0.93        27
          49       0.96      1.00      0.98        26
          50       1.00      0.88      0.94        17
          51       0.81      0.93      0.87        14
          52       0.00      0.00      0.00         5
          53       0.95      1.00      0.98        20
          54       0.94      0.84      0.89        19
          55       1.00      0.97      0.99        36
          56       0.85      0.85      0.85        20
          57       0.92      0.92      0.92        26
          58       1.00      0.88      0.94        25
          59       0.92      0.88      0.90        25
          60       0.95      0.95      0.95        39
          61       0.60      1.00      0.75         3
          62       0.00      0.00      0.00         5
          63       0.95      0.95      0.95        39
          64       0.93      0.93      0.93        14
          65       1.00      0.80      0.89        10
          66       1.00      0.86      0.92         7
          67       0.94      0.81      0.87        37
          68       0.91      0.88      0.90        49
          69       0.95      0.96      0.96        78
          70       0.89      0.97      0.93        35
          71       0.93      0.81      0.87        16
          72       0.96      1.00      0.98        46
          73       0.82      1.00      0.90        31
          74       0.93      1.00      0.97        14
          75       0.80      1.00      0.89         4

    accuracy                           0.92      1983
   macro avg       0.88      0.89      0.88      1983
weighted avg       0.92      0.92      0.92      1983


# KNN
knn = KNeighborsClassifier()
knn_param_grid = {'n_neighbors': [3, 5, 7]}
knn_grid = GridSearchCV(knn, knn_param_grid, cv=5)
knn_grid.fit(X_train, y_train)
knn_best = knn_grid.best_estimator_

knn_pred = knn_best.predict(X_test)
knn_accuracy = accuracy_score(y_test, knn_pred)
knn_precision = precision_score(y_test, knn_pred, average='weighted')
knn_recall = recall_score(y_test, knn_pred, average='weighted')
knn_f1 = f1_score(y_test, knn_pred, average='weighted')
knn_cm = confusion_matrix(y_test, knn_pred)
print("\nKNN Results:")
print("Accuracy: {:.3f}".format(knn_accuracy))
print("Precision: {:.3f}".format(knn_precision))
print("Recall: {:.3f}".format(knn_recall))
print("F1 Score: {:.3f}".format(knn_f1))

# Random Forest
rf = RandomForestClassifier()
rf_param_grid = {'n_estimators': [100, 200, 300], 'max_depth': [3, 5, 7]}
rf_grid = GridSearchCV(rf, rf_param_grid, cv=5)
rf_grid.fit(X_train, y_train)
rf_best = rf_grid.best_estimator_
rf_pred = rf_best.predict(X_test)
rf_accuracy = accuracy_score(y_test, rf_pred)
rf_precision = precision_score(y_test, rf_pred, average='weighted')
rf_recall = recall_score(y_test, rf_pred, average='weighted')
rf_f1 = f1_score(y_test, rf_pred, average='weighted')
rf_cm = confusion_matrix(y_test, rf_pred)
print("\nRandom Forest Results:")
print("Accuracy: {:.3f}".format(rf_accuracy))
print("Precision: {:.3f}".format(rf_precision))
print("Recall: {:.3f}".format(rf_recall))
print("F1 Score: {:.3f}".format(rf_f1))

KNN Results:
Accuracy: 0.921
Precision: 0.917
Recall: 0.921
F1 Score: 0.917

Random Forest Results:
Accuracy: 0.905
Precision: 0.892
Recall: 0.905
F1 Score: 0.896

	inm_floor	inm_size	inm_price	inm_longitude	inm_latitude	inm_barrio	inm_distrito	his_price	his_quarterly_variation	his_annual_variation	...	dem_TasaDeParo	dem_TamanoMedioDelHogar	dem_PropSinEstudiosUniversitarios	dem_PropSinEstudios	dem_Proporcion_de_nacidos_fuera_de_Espana	dem_PropConEstudiosUniversitarios	dem_PobTotal	dem_NumViviendas	dem_EdadMedia	dem_Densidad_(Habit/Ha)
0	3.0	141.0	990000	-3.656875	40.464347	Canillas	Hortaleza	3250	2.2	0.3	...	8.724674	2.527886	0.488949	0.175632	15.456193	NaN	40838	16155	NaN	161.894356
1	2.0	159.0	940000	-3.703523	40.419427	Universidad	Centro	5106	1.4	-4.3	...	9.006094	1.975877	0.386598	0.083812	32.102460	0.529590	33418	16913	43.678945	352.500616
2	NaN	NaN	549000	-3.669626	40.435362	Guindalera	Salamanca	4100	0.6	-4.1	...	7.441379	2.369951	0.365818	0.070351	18.224365	0.563831	42306	17851	46.477166	263.952286
3	2.0	232.0	750000	-3.720619	40.424164	Argüelles	Moncloa - Aravaca	4773	0.5	-3.7	...	6.709633	2.328217	0.343683	0.066403	20.963846	0.589914	24423	10490	46.972342	322.402577
4	4.0	183.0	1550000	-3.705909	40.413214	Sol	Centro	4739	-5.5	-5.3	...	9.058980	1.994244	0.433750	0.082242	39.490947	0.484009	7622	3822	44.632774	171.165183
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
17852	0.0	28.0	94000	-3.705236	40.448713	Bellas Vistas	Tetuán	3489	-0.3	-0.7	...	9.666273	2.483180	0.507461	0.139521	36.845626	0.353018	29895	12039	43.224268	418.012531
17853	1.0	247.0	1250000	-3.797123	40.458962	Aravaca	NaN	3683	3.1	0.8	...	5.072028	3.159086	0.318851	0.071317	15.663746	0.609832	27503	8706	40.112479	47.142277
17854	3.0	73.0	137260	-3.731940	40.377805	NaN	Carabanchel	1859	-0.9	-4.1	...	12.527023	2.858792	0.600245	0.221107	34.012583	0.178648	31947	11175	42.374887	204.190236
17855	5.0	72.0	304880	-3.704166	40.455556	Berruguete	Tetuán	3345	1.8	2.8	...	10.119032	2.476095	0.536667	0.153495	37.770383	0.309838	25843	10437	NaN	425.705154
17856	5.0	422.0	2490000	-3.678877	40.437514	Lista	Salamanca	5913	1.6	-0.5	...	6.249904	2.277080	0.335454	0.061745	22.502475	0.602801	21211	9315	45.826434	407.735677

	inm_floor	inm_size	inm_price	inm_longitude	inm_latitude	inm_barrio	inm_distrito	his_price	his_quarterly_variation	his_annual_variation	...	dem_TasaDeParo	dem_TamanoMedioDelHogar	dem_PropSinEstudiosUniversitarios	dem_PropSinEstudios	dem_Proporcion_de_nacidos_fuera_de_Espana	dem_PropConEstudiosUniversitarios	dem_PobTotal	dem_NumViviendas	dem_EdadMedia	dem_Densidad_(Habit/Ha)
0	3.0	141.0	990000	-3.656875	40.464347	Canillas	Hortaleza	3250	2.2	0.3	...	8.724674	2.527886	0.488949	0.175632	15.456193	NaN	40838	16155	NaN	161.894356
1	2.0	159.0	940000	-3.703523	40.419427	Universidad	Centro	5106	1.4	-4.3	...	9.006094	1.975877	0.386598	0.083812	32.102460	0.529590	33418	16913	43.678945	352.500616
2	NaN	NaN	549000	-3.669626	40.435362	Guindalera	Salamanca	4100	0.6	-4.1	...	7.441379	2.369951	0.365818	0.070351	18.224365	0.563831	42306	17851	46.477166	263.952286
3	2.0	232.0	750000	-3.720619	40.424164	Argüelles	Moncloa - Aravaca	4773	0.5	-3.7	...	6.709633	2.328217	0.343683	0.066403	20.963846	0.589914	24423	10490	46.972342	322.402577
4	4.0	183.0	1550000	-3.705909	40.413214	Sol	Centro	4739	-5.5	-5.3	...	9.058980	1.994244	0.433750	0.082242	39.490947	0.484009	7622	3822	44.632774	171.165183
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
17852	0.0	28.0	94000	-3.705236	40.448713	Bellas Vistas	Tetuán	3489	-0.3	-0.7	...	9.666273	2.483180	0.507461	0.139521	36.845626	0.353018	29895	12039	43.224268	418.012531
17853	1.0	247.0	1250000	-3.797123	40.458962	Aravaca	NaN	3683	3.1	0.8	...	5.072028	3.159086	0.318851	0.071317	15.663746	0.609832	27503	8706	40.112479	47.142277
17854	3.0	73.0	137260	-3.731940	40.377805	NaN	Carabanchel	1859	-0.9	-4.1	...	12.527023	2.858792	0.600245	0.221107	34.012583	0.178648	31947	11175	42.374887	204.190236
17855	5.0	72.0	304880	-3.704166	40.455556	Berruguete	Tetuán	3345	1.8	2.8	...	10.119032	2.476095	0.536667	0.153495	37.770383	0.309838	25843	10437	NaN	425.705154
17856	5.0	422.0	2490000	-3.678877	40.437514	Lista	Salamanca	5913	1.6	-0.5	...	6.249904	2.277080	0.335454	0.061745	22.502475	0.602801	21211	9315	45.826434	407.735677

	inm_floor	inm_size	inm_price	inm_longitude	inm_latitude	his_price	his_quarterly_variation	his_annual_variation	his_monthly_variation	dem_Indice_de_reemplazo_de_la_poblacion_activa	...	dem_TasaDeParo	dem_TamanoMedioDelHogar	dem_PropSinEstudiosUniversitarios	dem_PropSinEstudios	dem_Proporcion_de_nacidos_fuera_de_Espana	dem_PropConEstudiosUniversitarios	dem_PobTotal	dem_NumViviendas	dem_EdadMedia	dem_Densidad_(Habit/Ha)
count	14067.000000	16090.000000	1.785700e+04	17857.000000	17857.000000	17857.000000	17857.000000	17857.000000	17857.000000	17857.000000	...	17857.000000	17857.000000	17857.000000	17857.000000	17857.000000	16132.000000	17857.000000	17857.000000	16067.000000	17857.000000
mean	2.696453	149.013753	6.964702e+05	-3.688064	40.424817	3993.643389	0.437918	-1.814499	0.035549	58.387524	...	8.736896	2.487494	0.436518	0.124281	24.851753	0.439830	28281.995296	11515.008960	44.315032	233.372820
std	2.387135	160.485808	8.547274e+05	0.039164	0.031176	1589.895597	2.510048	3.847875	1.533065	23.241554	...	3.042722	0.300608	0.109068	0.076923	8.542049	0.181168	13117.299282	5322.072109	2.842113	122.653928
min	-1.000000	8.000000	1.600000e+04	-3.837697	40.328140	1290.000000	-16.100000	-11.600000	-5.200000	26.976744	...	3.643735	1.975877	0.246480	0.036422	8.884988	0.072421	2398.000000	872.000000	33.198916	2.267773
25%	1.000000	66.000000	1.899000e+05	-3.708170	40.402633	2578.000000	-1.100000	-4.300000	-0.600000	41.889843	...	6.291480	2.297491	0.340818	0.062095	18.746056	0.257582	18785.000000	8011.000000	43.161070	155.641754
50%	2.000000	98.000000	3.700000e+05	-3.692498	40.426586	4049.000000	0.500000	-2.200000	0.000000	51.910341	...	8.076921	2.466320	0.391468	0.085268	23.400147	0.512828	25023.000000	10637.000000	45.113343	228.165728
75%	4.000000	166.000000	8.490000e+05	-3.669516	40.446864	5106.000000	1.900000	-0.100000	0.900000	65.111111	...	10.802359	2.673612	0.551965	0.191202	31.219133	0.592631	36922.000000	15312.000000	46.244228	322.402577
max	60.000000	2400.000000	1.500000e+07	-3.545674	40.507659	8299.000000	12.200000	17.200000	8.100000	171.272727	...	18.030112	3.506552	0.641264	0.355659	53.296248	0.699977	67418.000000	26836.000000	49.332035	460.612389

	inm_floor	inm_size	inm_price	inm_longitude	inm_latitude	inm_barrio	inm_distrito	his_price	his_quarterly_variation	his_annual_variation	...	dem_TasaDeParo	dem_TamanoMedioDelHogar	dem_PropSinEstudiosUniversitarios	dem_PropSinEstudios	dem_Proporcion_de_nacidos_fuera_de_Espana	dem_PropConEstudiosUniversitarios	dem_PobTotal	dem_NumViviendas	dem_EdadMedia	dem_Densidad_(Habit/Ha)
0	3.00000	141.000000	990000	-3.656875	40.464347	Canillas	Hortaleza	3250	2.2	0.3	...	8.724674	2.527886	0.488949	0.175632	15.456193	0.439859	40838	16155	44.315387	161.894356
1	2.00000	159.000000	940000	-3.703523	40.419427	Universidad	Centro	5106	1.4	-4.3	...	9.006094	1.975877	0.386598	0.083812	32.102460	0.529590	33418	16913	43.678945	352.500616
2	2.69675	149.033403	549000	-3.669626	40.435362	Guindalera	Salamanca	4100	0.6	-4.1	...	7.441379	2.369951	0.365818	0.070351	18.224365	0.563831	42306	17851	46.477166	263.952286
3	2.00000	232.000000	750000	-3.720619	40.424164	Argüelles	Moncloa - Aravaca	4773	0.5	-3.7	...	6.709633	2.328217	0.343683	0.066403	20.963846	0.589914	24423	10490	46.972342	322.402577
4	4.00000	183.000000	1550000	-3.705909	40.413214	Sol	Centro	4739	-5.5	-5.3	...	9.058980	1.994244	0.433750	0.082242	39.490947	0.484009	7622	3822	44.632774	171.165183
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
17852	0.00000	28.000000	94000	-3.705236	40.448713	Bellas Vistas	Tetuán	3489	-0.3	-0.7	...	9.666273	2.483180	0.507461	0.139521	36.845626	0.353018	29895	12039	43.224268	418.012531
17853	1.00000	247.000000	1250000	-3.797123	40.458962	Aravaca	Tetuán	3683	3.1	0.8	...	5.072028	3.159086	0.318851	0.071317	15.663746	0.609832	27503	8706	40.112479	47.142277
17854	3.00000	73.000000	137260	-3.731940	40.377805	Aravaca	Carabanchel	1859	-0.9	-4.1	...	12.527023	2.858792	0.600245	0.221107	34.012583	0.178648	31947	11175	42.374887	204.190236
17855	5.00000	72.000000	304880	-3.704166	40.455556	Berruguete	Tetuán	3345	1.8	2.8	...	10.119032	2.476095	0.536667	0.153495	37.770383	0.309838	25843	10437	44.315387	425.705154
17856	5.00000	422.000000	2490000	-3.678877	40.437514	Lista	Salamanca	5913	1.6	-0.5	...	6.249904	2.277080	0.335454	0.061745	22.502475	0.602801	21211	9315	45.826434	407.735677

	inm_floor	inm_size	inm_price	inm_longitude	inm_latitude	inm_barrio	inm_distrito	his_price	his_quarterly_variation	his_annual_variation	his_monthly_variation	dem_Indice_de_reemplazo_de_la_poblacion_activa	dem_Indice_de_estructura_de_la_poblacion_activa	dem_Indice_de_dependencia	dem_Proporcion_de_nacidos_fuera_de_Espana	dem_PobTotal	dem_EdadMedia	dem_Densidad_(Habit/Ha)
0	0.458545	0.751935	1.835841	1.399435	1.523031	Canillas	Hortaleza	-0.534517	0.986104	0.829982	1.879833	-0.195840	-1.903755	1.628328	-1.362211	1.233897	-0.472593	-1.143597
1	-0.188578	1.113425	1.672118	-0.402840	-0.141066	Universidad	Centro	0.727408	0.597706	-0.694251	0.134325	-1.221461	1.907492	-2.055000	0.772796	0.498719	-0.900979	0.810418
2	0.262305	0.913268	0.391803	0.906798	0.449262	Guindalera	Salamanca	0.043412	0.209308	-0.627980	-0.127502	-0.138530	-0.204084	0.718097	-1.007172	1.379347	0.982490	-0.097342
3	-0.188578	2.579466	1.049970	-1.063352	0.034409	Argüelles	Moncloa - Aravaca	0.500996	0.160758	-0.495438	0.308876	-0.671695	0.047386	0.547483	-0.655814	-0.392511	1.315790	0.501865
6	0.458545	0.109287	0.034886	0.827278	-0.838365	Adelfas	Retiro	0.008736	-0.470389	-0.793658	0.047049	-0.325867	-1.605334	0.670161	-1.804264	-0.951127	0.122172	0.202023

Problem 7¶

Problem Description¶

First look at the data¶

Exploratory Data Analysis (EDA) & Data Cleaning¶

General Analysis¶

Missing Values¶

Univariate Analysis¶

Bivariate Analysis¶

Correlation Analysis¶

Scailing & Transformation¶

Splitting Dataset¶

Modeling¶

Classification Problem — Let's Predict Barrio!¶